As technologies have advanced—and especially with the rapid rise of AI that’s happening now—we’re seeing incredible capabilities and the most human-like computer intelligence we’ve ever encountered. Unfortunately, those capabilities take a big toll on the environment, sucking up huge amounts of energy and water and leaving a big carbon footprint that we’d be wise not to ignore. There are ways to reduce this negative impact, such as powering data centers in cool areas or training ML models at night. Using renewable energy is always a favorable idea, but in practice there is often a challenge to effectively harness the vast amounts of energy needed, and this work-in-progress still has a long way to go.
As is often the case with major problems, these efforts are certainly worthwhile but they come up short in chasing the pace and scale of the developing environment. In order to more fully solve the problem, we may need something different altogether. Something new and revolutionary—a turning point.
Recently I was fortunate to attend a talk by UC San Diego physics professor Dr. Alex Frañó. He and his team are researching possibilities for brain-like computing with quantum materials, which behave in oddly interesting ways under the right conditions. The team is exploring how the properties of these materials might lead to emergent behaviors similar to those that occur in the human brain. Such a phenomenon could lead to possibilities in neuromorphic computing, a bio-inspired approach that would greatly reduce the amount of energy needed in comparison to our current computing technologies.
Why Quantum Materials are so Cool
There are lots of ways quantum materials can make us scratch our heads in wonder, because they present us with paradoxes we’re not typically used to seeing or thinking about. For example, J.J. Thomson won the Nobel Prize in 1906 for his work demonstrating the behavior of electrons as particles. A few decades later, his son George won the Nobel Prize for showing electrons’ properties as waves. The father-son duo ended up pointing out an important oddity: electrons behave as both particles and waves. Light has the same behavior duality, seeming to travel in waves and also appearing as photon particles, depending on how you measure it.
Quantum materials are those atomic materials that can’t be described in terms of classical physics—they have different behaviors that appear through complex, conditional electronic states and can only be explained with quantum mechanics. Sometimes nothing much will occur with a material, until it reaches certain conditions (such as a particular temperature) and then it suddenly acts very differently. In his presentation, Dr. Frañó showed how a small block of a superconductor ceramic material seemed ordinary until he cooled it to a given temperature using dry ice. Once it reached that point, it was able to hover over a circular magnetic track, making its flight in rings until it eventually warmed up and succumbed to gravity like all the other objects on the table.
Emergence
Another interesting observation in physics is that when a bunch of tiny individual units join together, they can exhibit spontaneous phenomena that wouldn’t be expected by studying them individually—the whole becomes more than the sum of its parts. For a rough analogy, a conducted orchestra can create harmonies and a collective sound that wouldn’t be possible without a certain level of coordination. Otherwise, it would just be a bunch of individual instrumentalists doing their own thing. The fact that we can have orchestras is awesome, if you stop to think about it.
Emergence happens when the individual units interact in ways that influence one another, such that they begin to work together to form something larger and new. Although they’re following simple rules of physics, the ways that they organize together cause complex behaviors to arise. Those individual units could have never shown such behaviors on their own.
The brain shows a remarkable emergence itself: individual proteins make up billions of neurons which transfer electronic signals. Those signals influence one another, so that when they work together in sync, consciousness emerges. There are ways that quantum materials show behaviors similar to that of the brain, such as electronic spiking signal patterns, or the same kinds of reactional tendencies under changed parameters. Could these materials present an emergence like the brain’s, without needing a room full of servers to power them?
Past, Present and Future Technologies
Our human past involves key turning points when crucial discoveries were made that drastically sped up technological progress. (Check out this fascinating long-term timeline of technology for a cool visual that shows the accelerating pace of change.) These turning point moments have marked new understandings of materials and how to control them. It started with stone, as our ancestors learned to create tools and weapons with it. Later they began to control metal like bronze and iron. Each new ability to harness materials for our purposes led to great technological advancement.
In terms of computing, an important development was the use of transistors starting in the 1940s. Silicon became the material of choice due to its higher performance than other options. As Dr. Frañó sees it, we are currently in the Silicon Age, and greater understanding of how to make use of quantum materials could revolutionize our technological progress, ushering in the Quantum Age.
While certain conditions need to be met in order to harness the potential in these materials (like the lower temperature for the floating block I witnessed), their neuromorphic properties similar to what the brain shows could lead to the possibility of intelligent computing that’s orders of magnitude more energy efficient than what we’re doing now. Quantum materials could also prove to be very useful for energy storage and conversion in general too, all of which would be a game-changer for our current energy and climate crisis.
Getting Inspiration From the Brain
In seeking to use quantum materials for the development of AI, such as in pattern recognition or to create a chatbot like ChatGPT, the idea is not to reproduce the brain but rather to find inspiration in it. We can attempt to mimic certain qualities of the brain, ultimately leading to something that is not itself a brain but another mechanism with some similar capabilities—hopefully with energy consumption more like the brain’s and less like ChatGPT’s. If Dr. Frañó and other researchers in this field can get to that point, our technological timeline will take another turn, and we’ll open up the next era in humanity’s journey.